Chicago-Kent Law Review
Volume 79
Issue 3 Symposium: "Law &": Philosophical,
Psychological, Linguistic, and Biological Perspectives on
Legal Scholarship
Article 22
October 2004
Pernicious Ambiguity in Contracts and Statutes
Lawrence M. Solan
Follow this and additional works at: http://scholarship.kentlaw.iit.edu/cklawreview
Part of the Law Commons
Recommended Citation
Lawrence M. Solan, Pernicious Ambiguity in Contracts and Statutes, 79 Chi.-Kent. L. Rev. 859 (2004).
Available at: http://scholarship.kentlaw.iit.edu/cklawreview/vol79/iss3/22
This Article is brought to you for free and open access by Scholarly Commons @ IIT Chicago-Kent College of Law. It has been accepted for inclusion
in Chicago-Kent Law Review by an authorized editor of Scholarly Commons @ IIT Chicago-Kent College of Law. For more information, please
contact [email protected].
PERNICIOUS AMBIGUITY IN CONTRACTS AND STATUTES
LAWRENCE M. SOLAN*
INTRODUCTION
This Article explores an interpretive problem that is not adequately acknowledged by the legal system. I call it "pernicious ambiguity."' Pernicious ambiguity occurs when the various actors involved
in a dispute all believe a text to be clear, but assign different meanings
to it. Depending upon how the legal system handles this situation, a
case with pernicious ambiguity can easily become a crap shoot. If the
judge does not take heed of the competing interpretations as reflecting a lack of clarity, and if that judge happens to understand the
document in a way helpful to a particular party, that party wins.
Because the document is not seen as ambiguous, the document is
declared clear. In reality, however, the document is even less clear
than are ambiguous documents. The competing interpretations reflect
a complete communicative breakdown. If language worked so poorly
in general, then it would not be possible to have a language-driven
rule of law at all.
The problem, perhaps ironically, is that the concept of ambiguity
is itself perniciously ambiguous. People do not always use the term in
the same way, and the differences often appear to go unnoticed.
While all agree that ambiguity occurs when language is reasonably
susceptible to different interpretations, people seem to differ with
respect to whether those interpretations have to be available to a single person, or whether ambiguity occurs when different speakers of
the language do not understand a particular passage the same way. In
* Professor of Law and Director, Center for the Study of Law, Language and Cognition,
Brooklyn Law School. My thanks to Mark Baltin, Bob Freidin, Sam Glucksberg, Gillian
Grebler, and Charles Knapp for helpful discussion. I wish to thank Tara Lombardi. Marji
Molavi, Adrienne Oppenheim, and Jerry Steigman for their valuable contributions to this
project as research assistants. An earlier version of this paper was presented at the University of
Wisconsin, and I benefited from many comments there. Some of this research was conducted
while I was on sabbatical leave as a Visiting Fellow in the Department of Psychology at
Princeton University. The research was further supported by a summer research grant from
Brooklyn Law School.
L. 1call it that because John Darley came up with the expression. My thanks to him.
CHICAGO-KENT LA IV REVIEW
[Vol 79:859
addition, line drawing problems lead to disagreement about what
interpretations are reasonable. This Article will show how these different notions of ambiguity emerge, and offer some explanations
based on advances in linguistics, cognitive psychology, and the philosophy of language.
The concept of ambiguity is ambiguous in a second way that is
relevant to this Article. When discussing indeterminacy in meaning,
linguists and philosophers often distinguish between ambiguity and
vagueness. The former is used to describe situations in which an
expression can be understood in more than one distinct sense (e.g.,
river bank versus savings bank), while the latter refers to problems of
borderline cases (e.g., a piece of ceramic that is not clearly a bowl or a
cup, but something in between).' Legal writers, and judges in particular, use the word "ambiguity" to refer to all kinds of indeterminacy,
whatever their source. Because this Article focuses heavily on what
judges say, I will generally use the word ambiguity in this looser, legal
sense. However, as we will see below, the most pernicious cases of
pernicious ambiguity are actually instances of conceptual vagueness.'
Part I of the Article lays out the phenomenon of pernicious
ambiguity in more detail. Part II describes situations in which the
problem of pernicious ambiguity arises in both statutory and contractual interpretation. Courts are not of a single mind when it comes to
addressing this issue. Part III offers a linguistic explanation of how
the problem arises.
I.
COMMUNICATIVE FAILURE WITHOUT AMBIGUITY
Disagreement about meaning becomes pernicious if it is not discovered before something bad happens. When a party can recognize
ex ante that a misunderstanding is afoot, he can correct it before the
misunderstanding matures into a disagreement over obligation. The
Restatement (Second) of Contracts places an affirmative obligation
on the party recognizing the problem to tell the other party in ad-
2. This example comes from William Labov. See William Labov, The Boundaries of
Words and Their Meanings, in NEW WAYS OF ANALYZING VARIATION IN ENGLISH 349-51
(Charles-James N. Bailey & Roger W. Shuy eds., 1973). For discussion, see RAY JACKENDOFF,
SEMANTICS AND COGNITION 85-86 (1983).
3. For an accessible discussion of these distinctions by a linguist, see Sanford Schane,
Ambiguit) , and Misunderstanding ii the Law, 25 T. JEFFERSON L. REV. 167 (2002).
2(X4
1
PERNICIOUS AMBIGUITY
vance. If he does not, then the interpretation of the unwitting party
4
prevails.
But much of the time, the lack of communication continues until
undesired events occur, and the courts are then left to decide what to
do if they notice the problem at all. When this happens, the disagreement often manifests itself in the litigation positions of the parties,
whose interpretations may more realistically reflect their lawyers'
clever post hoc arguments than any serious dispute about meaning.
This possibility leads most courts to look with skepticism at differences in interpretation as evidence of ambiguity. But dishonesty
surely is not always the case, and courts that ignore disagreements
between parties in principle no doubt sell short some sincere litigants.
Part of the problem is that the law has only two ways to characterize the clarity of a legal text: It is either plain or it is ambiguous.
The determination is important. Whether the text is a statute, a contract, or an insurance policy, once a court finds the language to be
plain, it will typically refrain from engaging in a variety of contextually-based interpretive practices. For example, if a court declares the
language of a statute clear, it is unlikely to rely heavily on such things
as the history of the enactment process to determine whether the
statute should apply in a particular case. This 1917 statement by the
Supreme Court is typical:
It is elementary that the meaning of a statute must, in the first instance, be sought in the language in which the act is framed, and if
that is plain, and if the law is within the constitutional authority of
the law-making body which passed it, the sole function of the courts
is to enforce it according to its terms. 5
In criminal cases, an initial finding of ambiguity triggers the rule of
lenity-the principle that requires uncertainties in the meaning of a
law to be resolved in favor of the defendant.6
Similar principles apply to the law of contracts. If a contract is
clear, then resorting to extrinsic evidence that might undermine the
plain language of the agreement is barred by the parol evidence rule. 7
When, in contrast, contractual texts are deemed ambiguous, the reso4. Restatement (Second) of Contracts § 201 (1979).
5. Caminetti v. United States, 242 U.S. 470, 485 (1917).
6. See, e.g., Babbitt v. Sweet Home Chapter of Cmtys. for a Great Or., 515 U.S. 687, 704
n.18 (1995). 1 do not claim here that the rule of lenity is applied uniformly, only that it at least
potentially informs statutory interpretation in a subset of cases. See Lawrence M. Solan, Law,
Language, and Lenit'y, 40 Wm. & MARY L. REV. 57,92-94 (1998).
7. For a recent rationale, see W.W.W. Assocs., Inc. v. Giancontieri, 566 N.E.2d 639, 642
(N.Y. 1990).
CHICAGO-KENT LAW REVIEW
[Vol 79:859
lution of the ambiguity becomes a trial issue for the jury. Thus, a
court acts as a gatekeeper in making its initial inquiry into whether an
ambiguity exists. For this reason, whether to consider the disagreement of the parties or of other courts as evidence of ambiguity is an
important decision. It serves as a mechanism for choosing between
competing legal institutions, a choice that can affect the range of
possible outcomes of a dispute.8
Courts sometimes define ambiguity this way: "A statute is
ambiguous if it is susceptible of more than one reasonable interpretation." 9 The same description is used to describe contractual ambiguity: "In attempting to interpret such plans, our first task is to
determine if the contract at issue is ambiguous or unambiguous.
Contract language is ambiguous if it is susceptible to more than one
reasonable interpretation." 10 These statements are fair as far as they
go, but they leave unanswered an important question: Whose interpretation?
Typically, we say that language is ambiguous if we can interpret
it in more than one way. Linguists use the term in this sense.11 A classic example in the linguistic literature, from Noam Chomsky's early
work, is the sentence, "flying planes can be dangerous. ' '12 The sentence can refer either to planes aloft or to the act of piloting. The ambiguity arises from the fact that as speakers of English, we can assign
either of two structures to the sentence. Importantly, we all recognize
the availability of both interpretations. If the meaning of "flying
planes can be dangerous" were legally important, a judge would find
the sentence ambiguous and turn to contextual and other evidence.
Everyone would agree with at least that much, even if further legal
analysis were controversial.
But language can be indeterminate in another way. Assume that
you understand a text one way, I understand it another way, and neither of us finds it the least bit ambiguous. The text certainly is susceptible to more than one interpretation, but not by the same person.
8.
For discussion of the impact that institutional choice has on litigation, see NEIL K.
KOMESAR, LAW'S LIMITS (2001).
9. Rouse v. Iowa, 110 F. Supp. 2d 1117, 1125 (N.D. Iowa 2000).
10. Neuma, Inc. v. AMP, Inc., 259 F.3d 864, 873 (7th Cir. 2001) (internal citation omitted).
There are hundreds of cases that use this language to describe contractual ambiguity.
11. Linguists typically limit the concept of ambiguity to expressions that an individual can
recognize as having more than one interpretation. When different people interpret the same
language differently, linguists often speak of their having different "dialects" or different
"idiolects."
12. NOAM CHOMSKY, ASPECTS OF THE THEORY OF SYNTAX 21 (1965).
20041
PERNICIOUS AMBIGUITY
This is what happens when people understand the same language
differently, and don't have a chance to discover the problem. It is
made worse when the participants are all part of the same interpretative community, and assume that they speak and understand language
the same way. It is easy enough for me to figure out from only limited
experience that botanists call tomatoes fruit, and why they do so, even
though most of us persist in calling them vegetables.S Dialect differences, like the British boot and the American trunk of a car, may generate small initial confusion, but the context of the conversation
quickly saves the day. But when two people speak the same dialect,
their failure to understand each other is less likely to be detected. Yet
the legal community does not deal with this distinction systemati14
cally.
When this sort of communicative failure occurs in a legallyrelevant setting, it can be resolved at various junctures in the process:
1. During the formation process of a legal text. The law of contracts recognizes that people may understand the same language differently, and places the onus on the party who did notice, or should
have noticed, the miscommunication to straighten the matter out.1 5
Similarly, efforts are presumably made during the process of statutory
drafting to detect in advance language that may not be understood
uniformly by those to whom the statute is intended to apply.
2. After the language is in place, but before anything goes wrong.
This is most likely to happen when parties have enough contact with
each other to allow someone to discover the misunderstanding and to
convey it to the other party. One would expect this to occur in the
business context when the parties' ongoing relationship is a continuing one.1 6 The contact need not be friendly. For example, one party
may accuse the other of patent infringement. The accused party has
13. See Nix v. Hedden, 149 U.S. 304 (1893), where this issue actually arose in a customs
dispute. The Supreme Court sided with the ordinary language users over the experts, making
tomatoes, legally, a vegetable. Id. at 307. Since the tariff on vegetables at that time was higher
than that on fruit, the ordinary person won only the linguistic battle-not the economic oneover the price of produce. Id. at 305, 307.
14. These two perspectives on linguistic indeterminacy correspond to the distinction between "within subject" and "between subject" analysis used routinely by psychologists in analyzing data. This correspondence was pointed out to me by Bobbi Spellman, whom I thank.
15. Restatement (Second) of Contracts § 201 (1979).
16. Relevant here is the substantial literature on relational contracts. For an early discussion, see Stewart Macaulay, An Empirical View of Contract, 1985 Wis. L. REV. 465. For recent
discussion of relational contract theory, see Symposium, Relational Contract Theory: Unanswered Questions: A Symposium in Honor of Ian R. Macneil, 94 Nw. U. L. REV. 737 (2000), and
the many references cited in the papers in that volume.
CHICAGO-KENT LA IV REVIEW
[Vol 79:859
an opportunity to file a declaratory judgment action to determine the
parties' respective rights before much damage occurs. 17
3. After something bad has happened. Statutes and contracts are
both susceptible to this problem. As Peter Tiersma notes, legal documents don't talk back. 8 Potentially pernicious ambiguity in statutes
and other legal documents, therefore, is less likely to be resolved.
4. Never. Of course, some ambiguity is never discovered because
it never matters to anyone. When it does matter, people are likely to
be insensitive to diverging interpretations when commonly used
words play an important role in the legal system, and all players
assume that their own interpretation is the most natural. John Darley
and I have argued, based on experimental research we conducted,
that the concept of causation works this way. 19 It also may happen, at
least in part, when a judge instructs a jury. Research has shown that
jurors understand crimes in terms of mental models they have formed
in their minds as a result of their prototypes for crimes, and have
great difficulty eschewing these models in favor of the legal definitions upon which they are supposed to decide guilt and innocence. To
the legal system, an event may be a burglary. To the juror, it is not.2"
The same holds true with terms like reasonable doubt, and with the
term aggravation, which play an important role in death penalty
instructions, where it carries a meaning remote from how most people
use it in everyday speech."
Let us focus first on the third scenario: Disagreement over meaning comes to light, but only after something bad has happened.
Assume that two parties, P and D, offer different interpretations of a
document. Each is sincere, and each claims and honestly believes his
interpretation to be the only realistic one. Assume further that the
judge hearing the case understands the document the way P does.
That is, to the judge, the document is not susceptible "to more than
one reasonable interpretation." Perhaps the judge even gets angry at
D for trying to stretch the language unnaturally. When this happens,
17. See, e.g., Holmes Group, Inc. v. Vornado Air Circulation Sys., Inc., 535 U.S. 826 (2002).
18. Peter M. Tiersma, A Message in a Bottle: Text, Autonomy, and Statutory Interpretation,
76 TUL. L. REV. 431, 433 (2001).
19. Lawrence M. Solan & John M. Darley, Causation. Contribution,and Legal Liability:
An EmpiricalStud'y, 64 L. & CONTEMP. PROBS. 265 (2001).
20. See Vicki L. Smith, Prototypes in the Courtroom: Lay Representations of Legal Concepts, 61 J. PERSONALITY& SOC. PSYCHOL. 857 (1991).
21. See Peter Meijes Tiersma, Dictionariesand Death: Do CapitalJurors Understand Mitigation?, [995 UTAH L. REV. 1, 14-20, 43-47.
20041
PERNICIOUS AMBIGUITY
the system has equated pernicious ambiguity with clarity, a dubious
move at best.
Like the parties to a litigation, at times courts themselves may
not be sincere when they hold that the language of a statute is clear.
For example, a judge may believe that language is susceptible to a
number of interpretations, but say it is clear anyway in order to avoid
triggering an interpretive doctrine that would lead to a result that she
considers unjust in a particular case. 22 When interpretive doctrine
pushes judges toward putting more rhetorical weight on the language
than they may feel is just in a particular case, it would not be surprising to find that they write insincerely about language in order to reach
a result they believe is fair.
As for contracts, many jurisdictions follow the practice of looking
at extrinsic evidence to determine, as a preliminary matter, whether
the statute really is as clear as it might seem to be at superficial
inspection. If not, then the extrinsic evidence that was used to demonstrate the contract was not clear can be used substantively to show
that it should be interpreted inconsistently with the agreement's
seemingly plain language.2 3 Taken at face value, courts appear to
draw these lines with rigor and confidence. But using extrinsic evidence to detect ambiguity and using it to resolve ambiguity are not
sufficiently different activities to distinguish them in real life most of
the time.
Despite these (and other) problems, I believe there is value in
studying judicial arguments about ambiguity taken at face value. For
one thing, notwithstanding some inconsistency and perhaps even
some hypocrisy, courts do, I believe, take language arguments seriously. At the very least, discussion of the clarity of a legal text sets
boundaries for participants in the legal system. A lawyer might say,
"It's just rhetoric. I talk about clear and ambiguous language to
22. This happens often enough. See, e.g., Smith v. United States, 508 U.S. 223, 225 (1993)
(holding that because the language was clear, the rule of lenity should not be applied to a statute
enhancing sentences for using a firearm "during and in relation to [a] drug trafficking crime,"
when the defendant was convicted under the statute for attempting to trade a machine gun for
drugs). It also occurs regularly in the context of the Chevron doctrine, which requires courts to
defer to an administrative agency's interpretation of a statute unless the clear language of the
statute prohibits the agency's interpretation. For examples in the context of the interpretation of
statutes by administrative agencies, see Richard J. Pierce, Jr., Chevron and its Aftermath:
JudicialReview of Agency Interpretationsof Statutory Provisions,41 VAND. L. REV. 301 (1988).
23. See, e.g., Fed. Deposit Ins. Corp. v. W.R. Grace & Co., 877 F.2d 614,620 (7th Cir. 1989)
(Posner, J.) (using the terms "intrinsic" ambiguity and "extrinsic" ambiguity to capture the
notion that the determination of clarity must be made in context).
CHICAGO-KENT LA IV REVIEW
[Vol 79:859
please the judges, but I don't believe a word of it. ' ' 24 This may be so.
But that lawyer cannot get away with making arguments that are not
plausible. No lawyer representing a defendant in a bank robbery case
would argue that- the statute applies to picking flowers on the river
bank, or even more remotely, to regulations governing the disposal of
cadmium. Surely our understanding of language does some work in
limiting the range of plausible interpretations of legal texts, and a
great deal of work at that.
Perhaps more importantly, regardless of what judges say about it,
the problem of pernicious ambiguity has generated a body of mutually inconsistent judicial statements without much of an attempt to
resolve the incoherence in the body of cases. Whatever the ultimate
explanation, such phenomena often reflect deeper problems and are
worthy of investigation for that reason alone. The different approaches to ambiguity, whether result-oriented or not, would simply
not survive if we were not generally uncertain about what we mean
when we talk about ambiguity.
II. PERNICIOUS AMBIGUITY IN STATUTES AND CONTRACTS
This part of the Article examines some cases in which courts confront the problem that arises when different players in the legal
system interpret the same language differently, without acknowledging that there may be ambiguity. The cases discussed in Section A
illustrate how courts respond when parties take different positions
about meaning. Those described in Section B illustrate what happens
when judges themselves are on record as taking positions that are
inconsistent with each other. Section C examines some expressions
that experimental research suggests are understood differently by
different players in the legal system without any systematic recognition of the fact.
A.
When the PartiesDisagree
Courts are inconsistent in using disagreements between parties
about the meaning of a statute to draw inferences about the range of
reasonable interpretations that the statute affords. For example,
interpretation of the federal Bankruptcy Code is a great source of
24. For an interesting discussion of legal argument reduced to rhetoric, see Stanley Fish,
Dennis Martinez and the Uses of Theory, 96 YALE L.J. 1773 (1987). See also Pierre Schlag, No
Vehicles in the Park, 23 SEATrLE U. L. REV. 381 (1999).
20041
PERNICIOUSAMBIGUITY
disagreement among parties and judges. In one case that reached the
Supreme Court, Bank of America National Trust and Savings Ass'n v.
203 North LaSalle Street Partnership,25 the question involved the
rights of pre-bankruptcy shareholders to hold equity in a reorganized
corporation. It was a technical issue, which the majority resolved
against the old shareholders. Justice Thomas, joined by Justice Scalia,
concurred in the judgment, using the case as an opportunity to criticize the Court for its methodology in interpreting statutes both in that
case and in a then recent decision, Dewsnup v. Timm. 26
Section 506 was ambiguous, in the Court's view [in Dewsnup],
simply because the litigants and amici had offered competing interpretations of the statute. This is a remarkable and untenable methodology for interpreting any statute. If litigants' differing positions
demonstrate statutory ambiguity, it is hard to imagine how any
provision of the Code-or any other statute-would escape
Dewsnup's broad sweep. A mere disagreement among litigants over
the meaning of a statute does not prove ambiguity; it usually means
that one of the litigants is simply wrong. Dewsnup's approach to
statutory interpretation enables litigants to undermine the Code by
creating "ambiguous" statutory language and then cramming into
the Code any good idea that
can be garnered from pre-Code
practice or legislative history. 2
According to Justice Thomas, disagreement among the parties is evidence that one of the parties is wrong-not that there is an interpretive problem with the statute itself.
The Ninth Circuit has expressed this position succinctly:
"[S]tatutory ambiguity cannot be determined by referring to the
parties' interpretations of the statute. Of course their interpretations
differ. That is why they are in court. '28 Some states take this position
as well. For example, the Supreme Court of Idaho has held:
"Ambiguity
is not established merely because
differing
interpretations are presented to the Court; otherwise, all statutes
would be considered ambiguous." 29
Other courts come to the opposite conclusion, finding ambiguity
because of the parties' disagreement, provided that they each adopt
plausible positions. For example, a federal district court in Illinois was
confronted with the interpretation of a term in the Fair Labor Stan25. 526 U.S. 434 (1999).
26. Id. at 461 (Thomas, J., concurring) (discussing Dewsnup v. Timm, 502 U.S. 410 (1992)).
27. Id. (emphasis added).
28. John v. United States, 247 F.3d 1032. 1041 (9th Cir. 2001).
29. Inama v. Boise County, 63 P.3d 450 455 (Idaho 2003); see also BHA Investments, Inc.
v. City of Boise, 63 P.3d 482, 484 (Idaho 2003).
CHICAGO-KENT LA IV REVIEW
I Vol 79:859
dards Act. 30 The act requires that employers pay minimum wage to
employees, but lists some exceptions, including "companionship
services for individuals. ' 31 The question was whether the "homemaker services" provided by the defendant to elderly people in need
of assistance fit within that exception to the minimum wage law.
Although the court ultimately held that the exception did not apply, it
found the statute ambiguous:
Beginning with the statutory language, the plain meaning of § 215's
text is not sufficiently clear for this court to determine whether the
statute supports or contradicts DOL's regulation. As the parties'
plausible yet differing interpretations demonstrate, the term "companionship services" creates ambiguity in regard to the extent
to elderly
employees can provide household and personal3services
2
individuals without being consideredcompanions.
Similarly, the Supreme Court of Tennessee took seriously the
fact that the parties each took reasonable, but contrary positions in
interpreting the statute defining marital property in a divorce
proceeding. The court held:
This statute does not explicitly declare that benefits from a private
disability insurance contract are marital property, nor does the
statute specifically state that such benefits are separate property.
Where as here, the parties legitimately have different interpretations
of the same statutory language, an ambiguity exists, and we may
consider the legislative history and the entire statutory scheme for in-
terpretiveguidance.33
Notably, even the judges most generous in finding ambiguity
must first find that both interpretations are "plausible" or "legitimate" or, in other cases, "reasonable." But the only way that an
interpretation can meet this standard is if the judge deciding the question can comfortably see both interpretations as possible ones, given
the judge's everyday understanding of the language. Thus, even a
judge who believes that pernicious ambiguity should be taken seriously can take it only as seriously as her own language faculty allows.
Some courts adopt a "reasonably intelligent person" standard,
especially in the interpretation of insurance policies. An Indiana
appellate court wrote: "An insurance contract is ambiguous when it is
susceptible to more than one interpretation and reasonably intelligent
30. Harris v. Dorothy L. Sims Registry, No. 00-C-3028, 2001 WL 78448, at *1 (N.D. 111.Jan.
29, 2001).
31. 29 U.S.C. § 213(a)(15) (2000).
32. 2001 WL 78448 at *3 (emphasis added).
33. Gragg v. Gragg, 12 S.W.3d 412, 415 (Tenn. 200) (emphasis added).
2004]
PERNICIOUS AMBIGUITY
persons would honestly differ as to its meaning. An ambiguity does
not exist simply because a controversy exists between the parties,
with each favoring a different interpretation. '' 34 Massachusetts courts
take the same position.35 These cases generally conclude that the
disagreement between the parties is not evidence of ambiguity, implying that at least one of the parties (and the party's lawyer, who filed
the papers) is not a person of reasonable intelligence. The courts do
not acknowledge the possibility that a judge who does not see the full
range of interpretations that different people experience can easily
enough mistake subtle differences in dialect for a lack of intelligence.
Perhaps for reasons of politeness, these same courts do not adopt a
"reasonably intelligent judge" standard when the disagreement is
among various courts that have interpreted the same language differently.
Although courts interpreting statutes often concern themselves
with the intent of the legislature, they almost never concern themselves with the ex ante statutory interpretation of the person accused
of violating the statute, and certainly not with the government's state
of mind prior to an arrest in deciding whether a prosecution is valid.3 6
With contracts, however, the parties in the dispute presumably both
participated in the contract formation process. 37 Section 201 of the
Restatement, for example, holds that the intent of the parties should
prevail when they had identical intent; the intent of one party should
prevail when the other party knew or had reason to know that the
parties did not share a common understanding, but did nothing about
it; and there is no contract when there was no mutual understanding. 3
34. Stevenson v. Hamilton Mutual Ins. Co., 672 N.E.2d 467, 471 (Ind. Ct. App. 1996).
Although insurance policies are generally governed by the law of contracts, since most are not
negotiated in advance, it seems appropriate to discuss them along with statutory cases in this
regard.
35. See, e.g., Citation Ins. Co. v. Gomez, 688 N.E.2d 951,953 (Mass. 1998).
36. This is not always true, however. See Liparota v. United States, 471 U.S. 419 (1985);
Ratzlaf v. United States, 510 U.S. 135 (1994). The Supreme Court held in both cases that
Congress had intended knowledge of the law to be an element of the offense.
37. For interesting comparison of statutory and contract law concerning the use of extrinsic
evidence in interpretation, see Stephen F. Ross & Daniel Tranen, The Modern Parol Evidence
Rule and Its Implicationsfor New Textualist Statutory Interpretation,87 GEO. L.J. 195 (1998).
38. Section 201 of the Restatement (Second) of Contracts reads:
§ 201 Whose Meaning Prevails
(1) Where the parties have attached the same meaning to a promise or agreement or a
term thereof, it is interpreted in accordance with that meaning.
(2) Where the parties have attached different meanings to a promise or agreement or a
term thereof, it is interpreted in accordance with the meaning attached by one of them
if at the time the agreement was made
CHICAGO-KENT LA W REVIEW
I Vol 79:859
Still widely cited as an illustration of this approach is Raffles v.
Wichelhaus,39 an 1864 English case in which a seller of cotton agreed
to ship it on the ship Peerless. It turned out that there were two ships
with that name, and the buyer thought he was purchasing cotton that
was to be delivered on the ship that was to arrive earlier. The buyer
rejected the goods, and the court held that there was no contract
because the parties had not come to an agreement, each having legitimately understood the same language differently. Also illustrative
is Judge Friendly's opinion in Frigaliment Importing Co. v. B.N.S.
InternationalSales Corp.,40 which held that a contract for the sale of
chickens not to be enforceable when both parties reasonably believed41
that a different type of chicken was the subject of the agreement.
Significantly, it is only by looking at facts outside the language of the
contract itself that the legitimacy of a party's claim that there was no
mutual understanding can be evaluated.
Regardless of any seeming differences in doctrine between statutory and contractual interpretation, in practice the contract cases look
a great deal like the statutory cases. Courts ask whether the parties'
interpretations are "reasonable" as evidence of whether they actually
understood the contract ex ante as they say they did .4 To find out
what is reasonable, they look first to the language of the contract,
whose interpretation is subject to the parol evidence rule, and then
sometimes to extrinsic aids. In this context, courts frequently repeat
that "[a] contract is not rendered ambiguous simply because the parties do not agree on the meaning of its terms. '43 Moreover, courts are
loathe to find no agreement at all, and bend over backwards to attribute some mutual intent to the parties. The result is a formal
(a) that party did not know of any different meaning attached by the other, and
the other knew the meaning attached by the first party; or
(b) that party had no reason to know of any different meaning attached by the
other, and the other had reason to know the meaning attached by the first
party.
(3) Except as stated inthis Section, neither party is bound by the meaning attached by
the other, even though the result may be a failure of mutual assent.
39. 159 Eng. Rep. 375 (Ex. Ch. 1864). For recent discussion of this case in a judicial opinion, see, e.g., Rossetto v. Pabst Brewing Co., Inc., 217 F.3d 539, 543 (7th Cir. 2000) (Posner, J.).
40. 190 F. Supp. 116 (S.D.N.Y. 1960).
41. See Schane, supra note 3, for linguistic discussion of the nature of the ambiguity in both
of these cases.
42. See, e.g., Garber v. Provident Life and Accident Ins. Co., No. 98-3043, 1999 WL 357812,
at *1 (6th Cir. May 27, 1999); 1 E. ALLAN FARNSWORTH, FARNSWORTH ON CONTRACTS §
7.12(a) (2d ed. 1998).
43. Bourke v. Dun & Bradstreet Corp., 159 F.3d 1032, 11)36 (7th Cir. 1998).
PERNICIOUSAMBIGUITY
20041]
recognition of pernicious ambiguity, but it is not clear how much is
unacknowledged in practice.
B.
When the Courts Disagree
In Moskal v. United States, 4 the Supreme Court was faced with a
difficult problem of statutory interpretation. A statute bars the interstate transportation of "falsely made, forged, altered, or counterfeited
securities. ' 41 According to the statute, valid automobile titles count as
securities. 6 Moskal, who ran a business in Pennsylvania, was involved
in a "title washing" scheme. He would set back the odometers of used
cars he bought, record the false information on the appropriate
documents, and then receive clean titles from the state of Virginia.
The Virginia titles were genuine, but contained false information
about the car's mileage. At the time the statute was enacted, "falsely
made" was more or less a synonym for "counterfeit." Today, it
sounds more like a descriptive term meaning made to be false in some
way. By the time the Supreme Court addressed the issue in 1990,
many courts had ruled that "falsely made" means counterfeit.
Justice Marshall wrote the majority opinion, on behalf of six justices, affirming Moskal's conviction. He rejected the defendant's
argument that the statute must be at least ambiguous, triggering the
rule of lenity, based on the interpretation of lower courts:
Because the meaning of language is inherently contextual, we have
declined to deem a statute "ambiguous" for purposes of lenity
merely because it was possible to articulate a construction more
narrow than that urged by the Government. Nor have we deemed a
division of judicial authority automaticallysufficient to trigger lenity.
If that were sufficient, one court's unduly narrow reading of a criminal statute would become binding on all other courts, including this
one. Instead, we have always reserved lenity for those situations in
which a reasonable doubt persists about a statute's intended scope
even after resort to "the language and structure, legislative history,
and motivating policies" of the statute. 47
Thus, the Supreme Court took the view that differences of opinion by
judges about the meaning of a statute do not demonstrate that the
statute's language is subject to more than one reasonable interpretation. In fact, the majority decided that the statutory language was
44. 498 U.S. 103 (1990).
45. 18 U.S.C. § 2314 (2000).
46. Id. at § 2311.
47. Moskal. 498 U.S. at 465. (second emphasis added, internal citations omitted).
CHICAGO-KENT LA IV REVIEW
[Vol 79:859
clear, a conclusion that justified rejecting application of the rule of
lenity.
Other courts have followed suit. For example, the United States
Court of Appeals for the Fourth Circuit made the following statement
in finding supplemental jurisdiction in a products liability case involving a complicated array of parties, notwithstanding findings of other
courts to the contrary:
[W]e cannot allow the fact that other circuits have called a statute
ambiguous to negate this circuit's duty to interpret the text of the
enactment. To hold otherwise would mean that we would automatically call a statute ambiguous because a sister circuit has interpreted a statute in a contrary manner. In 48effect, we would be
abandoning our own duty to interpret the law.
And in a case concerning the rights of veterans to recover attorneys
fees in certain proceedings brought before the Court of Veterans Appeals, the United States Court of Appeals for the Federal Circuit
remarked: "[W]ere we to accept the Secretary's position that differences in judicial interpretation of a statute proved the statute's ambiguity, we could never reverse a court on plain-language grounds and
we would, in effect, be bound by any other court's construction."4 9
The interpretation of insurance policies is also subject to this
approach. A California court rejected an argument that different judicial interpretations of the expression "sudden and accidental" in
standard insurance policies demonstrates ambiguity. The court argued that different jurisdictions approach the question of ambiguity
differently, paving the way for inconsistent results. 5° Still other courts
have refrained from stating the issue as a general principle of interpretation, but have nonetheless not given weight to earlier judicial
interpretations as evidence of ambiguity. 51
A recurring issue in insurance law is whether closely related
events should count as one "occurrence" or as more than one occurrence. Most often, the issue arises in connection with whether the
insured will have to pay more than one deductible, since policies
typically include an amount per "occurrence" that the insured must
48. Rosmer v. Pfizer Inc., 263 F.3d 110, 118 (4th Cir. 2001).
49. Jones v. Brown, 41 F.3d 634,639 (Fed. Cir. t994); see also Olden v. LaFarge Corp., 203
F.R.D. 254, 262 (E.D. Mich. 2001) ("The fact that courts disagree on whether a statute is
ambiguous does not itself create ambiguity.").
50. ACL Techs., Inc. v. Northbrook Prop. & Cas. Ins. Co., 22 Cal. Rptr. 2d 206, 214 n.39
(Cal. Ct. App. 1993).
51. See, e.g., Lee v. Boeing Co., Inc., 123 F.3d 811t, 806 (4th Cir. 1997); United States v.
Wildes, 120 F.3d 468, 471 (4th Cir. 1997).
20041
PERNICIOUS AMBIGUITY
pay before the insurance company becomes responsible for payment. 52 In H. E. Butt Grocery Co. v. National Union Fire Insurance
Co.,53 an employee of the insured had sexually abused two children.
The insurance company said this should count as two occurrences; the
insured said it should count as one. Courts were divided with respect
to cases that were reasonably similar. In holding that the events were
two occurrences, the Sixth Circuit commented: "[U]nder Texas law,
even where courts from different jurisdictions are split as to the interpretation of a particular insurance provision, '[n]either conflicting
views of coverage, nor disputation is sufficient to create an ambiguity.'"54
Consider similarly the following example from Ohio, Park-Ohio
Industries v. Home Indemnity Co. 55 General commercial liability insurance policies often contain "pollution" exclusions. A company that
wants to obtain insurance against claims that it polluted must buy a
separate policy. The question in Park-Ohio Industries, and many
other cases around the country, is whether injuries caused by fumes
from a furnace in the building should be excluded as "pollution." The
company that made the furnace (i.e., the plaintiff) argued that the
pollution exception applied only if the insured was engaged in actively discharging pollutants, or discharging them on their premises.
The insurance company disagreed, and relied upon a rather broad
notion of pollution.
The plaintiff attempted to rely on the fact that other jurisdictions
had interpreted the policy provision in its favor. The argument was
that these favorable cases established at the very least that the policy
was ambiguous, thus triggering the long-established rule that ambiguities in insurance policies are to be resolved in favor of the insured.
An earlier Ohio appellate decision, Equitable Life Insurance Company of Iowa v. Gerwick,56 supported the plaintiff's position:
Where the language of a clause used in an insurance contract is
such that courts of numerous jurisdictions have found it necessary
52. But not always. The most significant case dealing with this issue is the litigation over
whether the destruction of the World Trade Center on September 11, 2001 should be considered one or two occurrences for purposes of construing an insurance policy for approximately
$3.5 billion per occurrence. See World Trade Center Props., L.L.C. v. Hartford Fire Ins. Co., 345
F.3d 154 (2d Cir. 2003).
53. 150 F.3d 526 (5th Cir. 1998).
54. Id. at 534 (quoting Union Pac. Resources v. Aetna Cas. & Sur., 894 S.W.2d 401, 401
(Tex. App. 1994) (emphasis added).
55. 975 F.2d 1215 (6th Cir. 1992).
56. 197 N.E. 923 (Ohio Ct. App. 1934).
CHICAGO-KENT LA IV REVIEW
I Vol 79:859
to construe it and in such construction have arrived at conflicting
conclusions as to the correct meaning, intent, and effect thereof, the
question whether such clause is ambiguous ceases to be an open
one. The rule in Ohio is that ambiguous language is to be construed
most strongly against the party selecting the language and is especially applicable to contracts executed subsequently to such conflicting constructions. Policies, which are prepared by the insurance
company, and which are reasonably open to different
57 interpretations, will be construed most favorably to the insured.
The Sixth Circuit rejected this argument as applied to the pollution exclusion as an abdication of its responsibility to apply the plain
language of the policy, which it believed supported the insurance
company's position. It held:
To the extent that Equitable Life Ins. Co. held that a contract provision is ambiguous as a matter of law if other jurisdictions have
interpreted similar provisions differently, that case, in our view,
does not state the present law in Ohio. The Supreme Court of Ohio
has held that "[wihen the language of an insurance policy has a
plain and ordinary meaning, it is unnecessary and impermissible for
this court to resort to constructions of that language." If we were to
accept plaintiffs' argument that a contract provision is ambiguous as
a matter of law because other jurisdictions have chosen to apply a
provision differently, then we would be rejecting a well-settled Ohio
of the contract where
rule of construction to apply the plain language
58
that language is clear and unambiguous.
Summary judgment in favor of the insurance company was affirmed.
With equal confidence, other courts come to precisely the opposite conclusion. The United States Court of Appeals for the First
Circuit, interpreting the Bankruptcy Code, noted the different positions taken by courts as evidence of ambiguity:
In our judgment, this collision of viewpoints underscores the obvious: although the text of section 365(a) plainly indicates that a trustee's rejection of a nonresidential lease is conditional upon court
approval, the text is unclear as to whether that approval constitutes
a condition precedent or subsequent to an effective rejection. Consequently, section 365(a) is ambiguous in this respect. 59
The D.C. Circuit accepted a similar argument:
Finally, in concluding that the Commission's position is not dictated
by the statute's text or legislative history, we observe that it would
be unusual for a statute free from ambiguity to be subject to
different interpretations by the Commission over time, or, more
57. Id. at 925.
58. Park-Ohio, 975 F.2d at 1220 (second emphasis added, internal citation omitted).
59. In re Thinking Machines Corp., 67 F.3d 1021, 1025 (1st Cir. 1995).
2(M I
PERNICIOUS AMBIGUITY
immediately,
by a closely divided panel in the decision under
60
review.
Along these same lines, a federal case from California involved
the interpretation of a term in the Comprehensive Environmental
Response, Compensation, and Liability Act of 1980 ("CERCLA"),61
a statute notorious for its bad drafting.6 2 In dealing with a thorny interpretive issue, the court noted: "The fact that a statute has been
interpreted differently by different courts is evidence that the statute
is ambiguous and unclear."63
Perhaps the Supreme Court of Wisconsin's tempered version of
this approach captures the main thrust of the position: "Differing
interpretations of a statute does not alone create ambiguity, but
rather equally sensible interpretations of a term by different authorities are indicative of a statute's ability to support more than one
meaning. ' 64 In fact, Wisconsin distinguishes between disagreements
among the parties, which it considers irrelevant, and disagreement
among "decision-making bodies," which it considers relevant in determining whether a statute is clear. 65
Standard insurance policy provisions trigger the same questions,
sometimes with inconsistent results. Consider the following passage
from a federal court in Texas:
Under Texas law, the Court must ascribe to policy terms their ordinary and generally accepted meaning, unless the policy shows that
the words were meant in a technical or different sense. Moreover,
policy ambiguities are to be resolved in favor of the insured. In the
instant case, the phrases "advertising ideas" and "style of doing
business" are not defined under the Travelers' policy. In light of the
disagreement among courts and scholars as to the meaning of these
terms, they can only be said to be ambiguous. Bay and FAE propose
an objectively reasonable interpretation of the policy language that
60. Local Union 1261 v. Fed. Mine Safety & Health Review Comm'n, 917 F.2d 42,46 (D.C.
Cir. 1990). Along these lines, the Fourth Circuit has held that differences of opinion within the
scholarly community provide strong evidence of ambiguity. Rehabilitation Ass'n of Va., Inc. v.
Kozlowski, 42 F.3d 1444, 1462-63 (4th Cir. 1994) ("Indeed, the fact that well-intentioned and
intelligent experts at legal exegesis have arrived at three or four seemingly plausible readings of
a particular text may be the best evidence that this interpretive puzzle has no definitive answer."
(Niemeyer, J., concurring in part, dissenting in part)).
61. 42 U.S.C. § 9601 et seq. (1989).
62. See John Copeland Nagle, CERCLA's Mistakes. 38 WM & MARY L. REV. 1405 (1997);
John Copeland Nagle, CERCLA, Causation,and Responsibility, 78 MINN. L. REV. 1493 (1994).
63. United States v. Iron Mountain Mines, Inc., 812 F. Supp. 1528, 1557 (E.D. Cal. 1992).
64. In re Paternity of Roberta Jo W., 578 N.W.2d 185, 189 (Wis. 1998) (internal quotations
omitted).
65. Stockbridge Sch. Dist. v. Dep't of Public Instruction Sch. Dist. Boundary Appeal Bd.,
550 N.W.2d 96,99 (Wis. 1996).
CHICAGO-KENT LA IV REVIEW
[Vol 79:859
has been accepted by several courts. Under Texas law, the Court is
reasonable interpretation most
required to adopt the objectively
66
favorable to the insured.
This statement is directly at odds with the statement to the contrary in
H.E. Butt Grocery Co.,67 discussed above, which was also decided by a
federal court under Texas law.
An insurance case decided by the Supreme Court of Wisconsin is
similarly inconsistent with that court's position on pernicious ambiguity in statutory cases. In Peace v. Northwestern National Insurance
Co.,68 the court decided that the pollution exclusion in a landlord's
liability policy applied when a tenant sued the landlord after having
suffered from lead poisoning caused by chipping leaded paint in an
apartment. As noted earlier, courts are divided on the interpretation
of pollution exclusion clauses. Only Chief Justice Abrahamson expressed concern about this division in interpretation as evidence of
ambiguity:
When numerous courts disagree about the meaning of language,
the language cannot be characterized as having a plain meaning.
Rather, the language is ambiguous; it is capable of being understood in two or more different senses by reasonably well-informed
might on careful analysis
persons even though one interpretation
69
seem more suitable to this court.
Courts really are of different minds when it comes to the issue of
trying to resolve pernicious ambiguity.
C.
When PerniciousAmbiguity Goes Unnoticed
Pernicious ambiguity also occurs with concepts that are part of
everyday legal discourse. These instances are more difficult to discover, and therefore, potentially more pernicious. Consider causation,
which plays such a large role in the law. John Darley and I presented
participants in a study with scenarios of situations in which the law
does not speak uniformly about whether liability should attach.7" One
scenario involved a person who left his key in the ignition of his car.
66. Bay Elec. Supply, Inc. v. Travelers Lloyds Ins. Co., 61 F. Supp. 2d 611, 617 (S.D. Tex.
1999) (emphasis added, internal citations omitted).
67. L50 F.3d 526 (5th Cir. 1998).
68. 596 N.W.2d 429 (Wis. 1999).
69. Id. at 449 (Abrahamson, C.J., dissenting).
70. Solan & Darley, supra note 19.
20041
PERNICIOUSAMBIGUITY
The car was later stolen by a teenager, who got into an accident.7
Should the key-leaver be held liable?
Among the questions we asked were ones concerning whether
the key-leaver should be considered a cause of the accident and/or
whether the key-leaver should be considered an enabler of the accident. The results are interesting. Some felt that the key-leaver both
caused and enabled the accident, and therefore found him liable.
Others felt that he did neither, and found him not liable. A third
group of subjects believed that the key-leaver enabled the accident to
occur, but didn't cause it. Within this third group, about half thought
that enablement was a good enough prerequisite for liability; the
other half would not hold the key-leaver liable.72
This means that not only do people have different judgments
about whether liability should attach in such situations, but they do
not share a single conceptualization of the entire situation. The same
held true for a second scenario which we presented to participants. A
social host (again, with varying states of mind) permitted an intoxicated guest to drive another guest home from a party. The driver got
into an accident, injuring the passenger. Should the host be liable?
Some said yes, some said no, just as courts and legislatures disagree
about the best answer to this question. But once again, our participants not only disagreed about the legal result, they disagreed about
the conceptualization of the situation. Some believed that either there
was both causation and enablement or neither causation nor enablement. Others believed that there was enablement, but not causation.7 3
All of this makes a difference, of course, to jurors who receive
instructions that they must find causation before they can impose
liability, and to judges ruling on defendants' motions to dismiss a case
because the plaintiff cannot establish causation. Certainly there are
political judgments to be made. But those judgments do not occur in a
vacuum. If, as matter of human cognitive capacity, people differ in
their understandings of these scenarios, then their conceptualizations
may lead them to reach different conclusions without ever recognizing that their analysis was not solely a result of applying a legal
standard-causation-to a set of facts. The voluminous case law on
71. We also manipulated the state of the key-leaver's mind and other factors relevant in
legal analysis.
72. Solan & Darley, supra note 9, at 289.
73. Id.
CHICAGO-KENT LA IV REVIEW
JVol 79:859
causation in tort cases generally shows very little recognition of this
problem.
A second example involves the concept of reasonable doubt. 7
Various courts define the concept differently, and experimental studies routinely show that it matters how the concept is defined. When
participants in a study are presented with a scenario, and different
participants receive different sets of instructions in which the definition of reasonable doubt is varied systematically, the percentage of
guilty verdicts differs according to the definition of reasonable doubt
in the instructions.
If reasonable doubt is defined by a court, and the jurors pay attention to the definition, then the issue is not one of pernicious ambiguity. There may be good reason to object to a particular definition
that a legal system employs, since some tend to shift the burden of
proof to the defendant, but at least the system has a uniform sense of
what it means by reasonable doubt. Pernicious ambiguity occurs when
a court determines that reasonable doubt is a concept well enough
understood by the community so that a definition is neither needed
nor appropriate. The United States Court of Appeals for the Seventh
Circuit, among others, has taken this position. 5 When this happens,
there is absolutely no way of knowing how different from each other
the jurors' concepts of reasonable doubt really are.
In fact, experimental evidence suggests that they are likely to be
quite different. An instruction promulgated by the Federal Judicial
Center, which associates the concept of reasonable doubt with proof
that leaves the jury "firmly convinced" of the defendant's guilt, appears best to result in convictions when the government's case is
strong, and acquittals when the government's case is weak.7 6 Other
74. For a summary of much of the empirical work on this concept, see Lawrence M. Solan,
Refocusing the Burden of Proof in Criminal Cases: Some Doubt About Reasonable Doubt, 78
TEX. L. REV. 105 (1999). For a recent linguistic approach to the concept using basic concepts,
see Anna Wierzbicka, 'Reasonable Man' and 'ReasonableDoubt': the English Language, Anglo
Culture and Anglo-American Law, 10 FORENSIC LINGUISTICS 1 (2003). For an insightful historical perspective on differing conceptions of reasonable doubt, see Steve Sheppard, The Metaphorphoses of Reasonable Doubt: How Changes in the Burden of Proof Have Weakened the
Presumption of Innocence, 78 NOTRE DAME L. REV. 1165 (2003).
75. See United States v. Blackburn, 992 F.2d 666 (7th Cir. 1993); United States v. Hall, 854
F.2d 1036 (7th Cir. 1988); United States v. Marquardt, 786 F.2d 771 (7th Cir. 1986); United
States v. Shaffner, 524 F.2d 1021 (7th Cir. 1975); United States v. Lawson, 507 F.2d 433 (7th Cir.
1974).
76. See Irwin A. Horowitz & Laird C. Kirkpatrick, A Concept in Search of a Definition: The
Effects of Reasonable Doubt Instructions on Certainty of Guilt Standards and Jury Verdicts, 20
LAW & HUM. BEHAV. 655 (1996); this and other studies are discussed in Solan, supra note 74.
2(9)41
PERNICIOUS AMBIGUITY
approaches, especially ones which focus on the kinds of doubts that
jurors should ignore, often lead to high conviction rates even when
the cases are weak, according to the experimental literature. At least
before instructions help guide their thinking, it appears that jurors
enter the system with differing concepts of reasonable doubt. There is
no way of detecting what those concepts are and no way of knowing
what effect these differences may have on the outcome of a case.
To take another example, research by Vicki Smith77 suggests that
jurors' everyday understanding of the definitions of crimes more than
likely plays a role in their deliberations, pushing aside the legal definitions that are read to them in instructions. Smith asked subjects to list
"features that were common to the category in general" for several
crimes.78 The most prevalent attributes of kidnapping, for example,
were "ransom demand," "victim is a child," and "victim is taken
away." The legal definition, "defendant secretly confines victim
against his/her will, or takes/forces/entices victim to another place
with intent to secretly confine,'79 contains features that everyday understanding does not, and lacks features that everyday understanding
contains. Smith presented participants with stories that varied systematically according to whether all the legal elements of kidnapping
were met, and whether the most salient features were present. The
results Were that it is very difficult to convince people that the legal
definitions should trump their notions of the crime based on their
prototypical experiences with the concepts.
Examples of this phenomenon are, obviously, not easy to enumerate because, by their very nature, different understandings of the
same concept tend to go unnoticed. Put simply, undiscovered pernicious ambiguity is hard to discover. One would expect that contracts,
especially complex contracts, are a rich source of these kinds of misunderstandings. That they do not destroy legal relationships suggests
that our use of language is sufficiently uniform to allow us to proceed
as if there were no problems at all, and to take our chances with
communicative failures. The next Part speculates about why this is
not a bad bet.
77. See Smith. supra note 20.
78. Id. at 860.
79. Id.at 861.
CHICAGO-KENT LA IV REVIEW
[Vol 79:859
III. How PERNICIOUS AMBIGUITY ARISES
The question of pernicious ambiguity can be seen as a special
case of a problem that philosophers have raised for years: How can
you tell whether a person seemingly speaking your language is following the same rules that you do when you speak your language? Of
course, at some level, the answer is that you can't. We all take our
chances. Sometimes the problem is easy enough to discover. When I
speak with a person from Boston who says "pop" where I would say
"soda," I can make an adjustment. I also may keep my guard up a
little more, knowing that dialect differences may lead to some trouble
in communication. But the basic assumption under which we all operate is that a person who speaks my language somehow "has" a system
for speaking and understanding my language in her head, whatever
that requires, and uses it more or less the same way I do. Chomsky
puts it this way:
It may be that when he listens to Mary speak, Peter proceeds by
assuming that she is identical to him, modulo M, some array of
modifications that he must work out. Sometimes the task is easy,
sometimes hard, sometimes hopeless. To work out M, Peter will use
any artifice available to him, though much of the process is doubtless automatic and unreflective.8 °
There is nothing remarkable about this assumption. We make
frequent assumptions that others share aspects of our human condition. Some are physical. We are not surprised to see people wearing
heavy clothes on a cold day. We are surprised to see them wearing
heavy clothes on a hot day. We are not surprised to see people eat.
And we expect, without evidence, that people have similar respiratory systems and need air in more or less the same way that we do,
that public buildings will necessarily have toilets, and so on. We make
similar assumptions about people's cognitive make up. If you stay at
my house and I tell you that the blue towel is yours, I expect (barring
color blindness) that you will not think I meant the brown one. I
expect that you perceive colors the same way I do and that we share
similar color vocabularies to describe these shared perceptions. The
same goes for shapes, patterns and many other things, including, crucially, what I mean when I use the words I do in conversation.
At this point, a difficulty in terminology arises, which begins to
account for why we have pernicious ambiguity. Whether you decide
80.
(2000).
NOAM
CHOMSKY,
NEW HORIZONS
IN THE STUDY OF LANGUAGE
AND MIND
30
20041
PERNICIOUS A MB IGUITY
to wear a warm coat is a matter of certain facts about the world and
certain facts about your state of mind. The coat, of course, is an
object, but the decision is solely a mental event. Similarly, when I tell
you to use the blue towel, it seems that the facts are both about you
and about the actual color of the towel." Even if every word I say is
mediated by my mental impressions of things out there, I seem to be
using language to refer to something in the world. If you choose the
brown towel instead of the blue one, I might think that you are rude,
or forgetful.
When I say, "you and I speak English," it appears from the way I
put it that there is something out there called English (just as there
are coats and towels out there), and you and I both speak it. It's convenient enough to put it that way, but speaking English, unlike using
a towel, is entirely about the person who is said to speak or understand it. If I say about a visiting foreign professor at my institution
that his English is serviceable, but not perfect, I am making a judgment about his knowledge of such things as pronunciation, syntax,
word usage, and the size of his vocabulary, and I am saying that his
state of knowledge is incomplete-or perhaps contains misinformation-compared to mine and compared to what I believe to be the
state of knowledge of other fluent speakers of the language."
Generally, unless I am provided with clues, such as obvious differences in dialect, or incomplete knowledge of a foreign language, or
the suspicion that we are somehow "talking past each other" in conversation, I default to the strategy of assuming that your mental state
and mine are the same with respect to the language we speak. This is
confirmed and reconfirmed every second we interact. If I say, "my car
is in the repair shop," I believe that you know exactly what I mean by
this, and my belief is most probably well-founded.
81. There is a great deal of debate about whether, when we use words that apparently refer
to things in the world, we are referring to an actual thing, our conceptualization of that thing, or
both. In one sense, we cannot refer to anything other than our conceptualization of a thing.
That's all we have. Even if this is all we refer to, however, we do so with confidence that our
conceptualization corresponds well enough to some kind of external reality to make our reference useful. For a recent discussion, see RAY JACKENDOFF, FOUNDATIONS OF LANGUAGE:
BRAIN, MEANING, GRAMMAR, EVOLUTION 300-06 (2002).
82. Chomsky refers to the personal nature of knowledge of language as "I-Language,"
meaning "internal language," as opposed to "E-Language," or "external language." CHOMSKY,
snpra note 80; NOAM CI-OMSKY, KNOWLEDGE OF LANGUAGE: ITS NATURE, ORIGIN, AND
USE (1986). There is considerable debate about these issues in the philosophical literature. For a
perspective that speaks in terms of language as an external phenomenon, see DONALD
DAVIDSON, INOUIRIES INTO TRUTH AND INTERPRETATION (1984).
CHICA GO-KENT LA IV REVIEW
[Vol 79:859
Of course, these assumptions can be completely wrong, even in
the most basic cases. Quine, for example, raised the question in the
context of what he called "radical translation. 8 3 Quine imagines a
linguist doing field work, trying to learn the language of a person in a
distant place, knowing nothing in advance about the language or the
culture. His only evidence comes from matching the informant's
words to events perceived in the physical world. The two see a rabbit
running, and the informant says, "Gavagai." The linguist notes that
' 4
"Gavagai" is the word for rabbit, or for, "lo, a rabbit. 8 Quine then
asks skeptically:
Who knows but what the objects to which this term applies are not
rabbits after all, but mere stages, or brief temporal segments, of
rabbits? In either event the stimulus situations that prompt assent
to 'Gavagai' would be the same as for 'Rabbit.' Or perhaps the objects to which 'gavagai' applies are all and sundry undetached parts
of rabbits; again the stimulus meaning would register no difference.
When from the sameness of stimulus meanings of 'Gavagai' and
'Rabbit' the linguist leaps to the conclusion that a gavagai is a
whole enduring rabbit, he is just taking for granted that the native
is enough like us to have a brief general term for rabbits and no
brief general term for rabbit stages or parts.8 5
Quine is right about these observations, but nothing changes as a
result. Indeed, if I were the linguist conducting the fieldwork, I would
do exactly what Quine's linguist did. If I turned out to be lucky, I
would ultimately come to a fairly accurate assessment of my informant's state of mind concerning Gavagai. If not, I would remain mistaken. In addition, as Chomsky notes, I would look at other sources
of information, beyond the physical world.8 6 Work in cognitive psychology suggesting that people have a bias toward interpreting what
they perceive as whole objects may shed some light on how87 likely I
am to be correct if I assume that the informant means rabbit.
This gives me added comfort in relying on the default assumption
that my informant, at least with respect to rabbits, interprets what he
sees the same way I do. In other words, I don't have to be quite as
lucky as Quine assumes. It also explains the ambiguity in ambiguity.
When I interpret failures in the default assumption as resulting from
83.
WILLARD VAN ORMAN QUINE, WORD AND OBJEcT
84. Id. at 29.
85. Id. at 51-52.
86. CHOMSKY, supra note SO, at 52-55.
87. See, e.g., PAUL BLOOM, HOW CHILDREN
ELLEN
M.
MARKMAN,
INDUcTION (1989).
CATEGORIZATION
AND
26 (1960).
LEARN THE MEANINGS OF WORDS (2000);
NAMING IN CHILDREN: PROBLEMS OF
20041
PERNICIOUSAMBIGUITY
differences between us, I am comfortable saying that an expression is
reasonably subject to more than one interpretation, and therefore,
ambiguous. On the other hand, when I interpret failures in the default
assumptions as your inability or refusal to identify an object in the
world properly, then I am less likely to attribute this misunderstanding to ambiguity. If I do not notice that I vacillate between tolerance
and intolerance of others in this respect, then the ambiguity of ambiguity has become pernicious.
Of course, misunderstandings occur all the time in less exotic
circumstances when we interact with people from other cultures. 88
Anna Wierzbicka has conducted an insightful study of emotion terms
across languages.89 Almost none are universal, but many of them
translate closely enough so that misunderstanding may be more the
rule than the exception. In fact, it's hard enough for many to communicate well about emotions using the vocabulary of their own language to other native speakers.
Why do these failures in communication occur? As an initial
matter, the problem appears to cut across traditional distinctions
between vagueness and ambiguity. Cases like the Peerless case are
about ambiguity of reference, while cases like those concerning what
constitutes pollution are about borderline cases, the hallmark of
vagueness. 90 But cases of actual ambiguity are less likely to remain
pernicious once the problem is brought out into the open, even if that
does not happen until a dispute arises. It would be strange indeed for
a judge deciding the Peerless case to say that the contract was clear
despite proof that the parties were referring to two different ships
with the same name. The more difficult cases are about disagreement
over the status of borderline cases. Below I offer a tentative explanation for how these disagreements arise.
88. How often they occur is a matter of debate. In his very interesting book, Talbot J.
Taylor presents a skeptical approach to understanding that allows for misunderstanding to be
rampant, and discusses various theories of language in light of this possibility. TALBOT J.
TAYLOR, MUTUAl
MISUNDERSTANDING: SCEPTICISM AND THE THEORIZING OF LANGUAGE
AND INTERPRETATION (1992). As discussed infra in the conclusion to this Article, I believe that
the problem of misunderstanding is concentrated in specific areas of linguistic knowledge, and
subject to analysis.
89. ANNA WIERZBICKA, EMOTIONS ACROSS LANGUAGES AND CULTURES: DIVERSITY
AND UNIVERSALS (1999).
90. For discussion of vagueness as a concept, see ROY SORENSEN, VAGUENESS AND
CONTRADIC'ION (2001). For discussion of differences between ambiguity and reference in legal
contexts, see Solan, supra note 74.
CHICAGO-KENT LA IV REVIEW
IVol 79:859
Let us return to the example of causation. Darley and I found
that people disagree about whether certain instances of enablement
fall within their concept of causation, or whether enablement is conceptually different. But if I say, "the bank robber murdered the teller
by putting two bullets through her head," there would be complete
consensus that the bank robber caused the death of the teller. The
tension between causation and enablement doesn't arise there. In
fact, it usually doesn't arise. It is precisely because we so often can say
with confidence that we know causation when we see it that makes us
think that our conceptualization of causation in the hard cases is the
only sensible one. Usually, we do know causation when we see it, and
when we do, our conceptualization is reinforced when others appear
to understand it the same way we do.
This phenomenon is rather typical. Experimental work by
Eleanor Rosch in the 1970s showed that people are very good at identifying prototypical instances of a category. 9' If I ask you how good an
example a chair is of the category furniture, we will all agree that it is
a very good example. But once we get into iffy cases, consensus
begins to dissipate. If I ask you to rank on a 1 to 7 scale how good an
example a piano is of the category furniture, where "1" is an excellent
example, "4" is "I don't know whether I'd call a piano furniture" and
"7" is "I'm certain that a piano is not an example of furniture," we
will get very few l's and 2's, but the answers are likely to span most of
the rest of the range.
This means that to the extent that we tend to use words in their
prototypical sense (consistent with the "ordinary meaning" canon of
statutory interpretation and contract law),92 we have good reason to
believe that others share our categorization decisions. But as we stray
from prototypical situations into hard cases, not only do we become
less certain about our own judgments, but consensus dissipates without our having any way of noticing the deterioration in common
understanding, unless it is brought to our attention. This, I believe, is
at the core of pernicious ambiguity in most of the cases in which it
arises in legal interpretation.
91. See Eleanor
Rosch, Cognitive Representations of Semantic Categories, 104 J.
EXPERIMENTAL PSYCHOL. 192 (1975).
92. The canon is routinely cited in judicial decisions. See, e.g., Smith v. United States, 508
U.S. 223, 228 (1993) ("When a word is not defined by statute, we normally construe it in accord
with its ordinary or natural meaning."). For discussion of the assumptions about language
underlying this canon, see Lawrence M. Solan, Finding Ordinary Meaning in the Dictionary', in
LANGUAGE AND LAW: PROCEEDINGS OF A CONFERENCE (M. Robinson ed., 2003).
PERNICIOUS AMBIGUITY
2004]
To take a more legally relevant example, consider the concept
lying, which plays a role in the law of perjury. In an important article,
linguists Linda Coleman and Paul Kay hypothesized that lying is a
family resemblance category.93 Three features contribute to our calling a statement a lie: actual falsity, intention to deceive, and whether
the speaker believes the statement to be false. When all three of these
features are present, we are most likely to call a statement a lie. When
none is present, we will not call it a lie. When one or two are present
in various combinations, we equivocate. Coleman and Kay presented
people with eight stories, covering all the possible combinations of
these features, and asked them to rate, on a I to 7 scale, whether they
thought the person lied, and if so, how sure they were that the person
lied. A "1" indicated "very sure it was not a lie," while "7" indicated
''very sure it was a lie." A "4" indicated "can't say." Consider the
following Clintonesque vignette taken from their article:
John and Mary have recently started going together. Valentino is
Mary's ex-boyfriend. One evening John asks Mary, 'Have you seen
Valentino this week?' Mary answers, 'Valentino's been sick with
mononucleosis for the past two weeks.' Valentino has in fact been
sick with mononucleosis for the past two weeks, but it is also the
case that 94Mary had a date with Valentino the night before. Did
Mary lie?
The mean score for this story was 3.48, near the midpoint of 4.95 But
this does not mean that everyone equivocated. They didn't. While the
precise distribution of scores is not available, Coleman and Kay
report that eighteen subjects thought it was a lie, seven couldn't say,
and forty-two said it was not a lie. 96 I have presented this vignette to
students on several occasions, and invariably get a wide range of reactions. Apparently, Mary would have won her impeachment trial also,
at least if she got the right judge. 97
Consider another vignette:
93. Linda Coleman & Paul Kay, Prototype Semantics: The English Word Lie, 57
LANGUAGE 26 (1981).
94.
95.
96.
97.
Id. at 31.
Id. at 33.
Id. at 39.
For discussion of the Clinton scandal in the light of Coleman and Kay's work, see
LAWRENCE M. SOLAN &
PETER M.
TIERSMA, SPEAKING OF CRIME:
THE LANGUAGE OF
CRIMINAL JUSTICE (forthcoming); Lawrence M. Solan, Perjury and Impeachment: The Rule of
Law or the Rule of Lawyers?, in AFTERMATH: THE CLINTON IMPEACHMENT AND THE
PRESIDENCY IN THE AGE OF POLITICAL SPECTACLE (Leonard V. Kaplan & Beverly I. Moran
eds., 2001); STEVEN L. WINTER, A CLEARING IN THE FOREST: LAW, LIFE, AND MIND (2001);
Peter Tiersma, Did Clinton Lie?: Defining "Sexual Relations", 79 CHI.-KENT L. REV. 927
(2004).
CHICAGO-KENT LA W REVIEW
I[Vol 79:859
Superfan has got tickets for the championship game and is very
proud of them. He shows them to his boss, who says, 'Listen, Superfan, any day you don't come to work, you better have a better
excuse than that.' Superfan says, 'I will.' On the day of the game,
Superfan calls in and says, 'I can't come to work today, Boss, because I'm sick.' Ironically, Superfan doesn't get to go to the game
because the slight stomach ache he felt on arising turns out to be
sick when he said he
ptomaine poisoning. So
98 Superfan was really
was. Did Superfan lie?
This story rated a mean of 4.61, just on the wrong side of lying. 99
Thirty-eight subjects called it a lie, twenty-one said it wasn't a lie, and
eight couldn't say. 100
Thus, it is not the case that there are only lies and truths. There
are also gray areas. But the gray areas do not consist only of instances
that we would all recognize as borderline cases. They also contain
many cases upon which people might disagree. Those who have
strong feelings about each of the two vignettes discussed above are
likely to have no real reason to know in advance that others may
conceptualize these stories differently. Perhaps more disturbing,
when presented with the views of others, we often think they are simply wrong, or even dishonest. This happens, I believe, because of our
mistaken conception of language as an external phenomenon, and the
deterioration of consensus about the status of nonprototypical
categorical decisions.
In the resolution of legal disputes, this phenomenon introduces a
great deal of chance into the system. Consider, for example, United
States v. DeZarn,10 1 a perjury case decided by the Sixth Circuit in 1998.
The case involved a general in the Kentucky National Guard improperly holding political fundraisers in which officers were expected to
contribute. One of those officers was DeZarn. He was asked the following questions under oath as part of the investigation:
Q: Okay. In 1991, and I recognize this is in the period that you were
retired, he held the Preakness party at his home. Were you aware
of that?
A: Yes....
Q: Okay. Sir, was that a political fundraising activity?
98.
99.
100.
101.
perjury
Coleman & Kay, supra note 93, at 31-32.
Id. at 33.
Id. at 39.
157 F.3d 1042 (6th Cir. 1998). Peter Tiersma and I discuss this case in detail in the
chapter of our forthcoming book. See SOLAN & TIERSMA, supra note 97, at ch. I1.
2004]
PERNICIOUS AMBIGUITY
A. Absolutely not.10 2
The Preakness Party really occurred in 1990. There was no
Preakness party in 1991. The questioner had made a mistake. DeZarn
defended himself by arguing that he had spoken truthfully. Nonetheless, he was prosecuted for perjury and convicted. The Sixth Circuit
affirmed. Like Superfan, the fact that his answer turned out to be
literally true did not save him. His intent to deceive carried the day.
If I am right about people differing about what they would call a
lie, then I would expect some disagreement about the conceptualization of this case, even if most people think the decision was appropriate. To the extent this is so, it suggests that the result of the case
hinges at least in part on who the judges are. The risk of arbitrariness
increases further when only a single judge decides, since in a panel of
three, there is at least some increased opportunity for pernicious
ambiguity to be discovered. But it may not be discovered there either.
CONCLUSION
Pernicious ambiguity is here to stay, and it is a challenge to rule
of law ideals. Yet it does not seem to arise so often in legal cases as to
be much of a challenge. The reason for this, I believe, is that, as discussed earlier, most cases of pernicious ambiguity are conceptual in
nature. They involve disagreements about whether a particular thing
or event is a borderline member, a clear member, or a nonmember of
a larger category. They rarely involve types of failures of communication, such as ambiguity of reference or syntactic ambiguity. Thus, a
great deal of what we do to understand language seems to occur
without producing the problem. It is not that there are no ambiguities
in these domains. Rather, what ambiguity there is in these domains is
easy enough to identify once the problem is pointed out, so that misunderstandings are less likely to be sustained. Thus, a great deal of
authoritative legal language goes undisputed. Moreover, since the
cases discussed in this Article are, at best, nonprototypical instances
of categories that have many clear cases, it should not be surprising
that the clear, prototypical cases dominate the legal landscape.
Can anything be done about pernicious ambiguity? Courts are
rightly cautious about drawing sweeping conclusions about uncertainty merely from the fact that there is a dispute over the applicability of language in a contract or a statute. Yet they are probably too
102. DeZarn, t57 F.3d at 1045.
888
CHICAGO-KENT LAW REVIEW
[Vol 79:859
cautious. When it appears that different people-especially randomly
selected judges with less of an agenda than litigation parties-are not
in accord about the interpretation of statutory or contractual language, it should raise a red flag that there may be a failure to communicate. Courts may be right in refusing to use these disagreements as
per se evidence of ambiguity. But they are dead wrong to conclude
that language upon whose meaning there is total disagreement is clear
language. They should stop doing so.